Landslide inventory mapping (LIM) is a key prerequisite for landslide susceptibility evaluation and disaster mitigation. It aims to record the location, size, and extent of landslides in each map scale. Machine learning algorithms, such as support vector machine (SVM) and random forest (RF), have been increasingly applied to landslide detection using remote sensing images in recent decades. However, their limitations have impeded their wide application. Furthermore, despite the widespread use of deep learning algorithms in remote sensing, for LIM, deep learning algorithms are limited to less unbalanced landslide samples. To this end, in this study, full convolution networks with focus loss (FCN-FL) were adopted to map historical landslides in regions with imbalanced samples using an improved symmetrically connected full convolution network and focus loss function to increase the feature level and reduce the contribution of the background loss value. In addition, K-fold cross-validation training models (FCN-FLK) were used to improve data utilization and model robustness. Results showed that the recall rate, F1-score, and mIoU of the model were improved by 0.08, 0.09, and 0.15, respectively, compared to FCN. It also demonstrated advantages over U-Net and SegNet. The results prove that the method proposed in this study can solve the problem of imbalanced sample in landslide inventory mapping. This research provides a reference for addressing imbalanced samples in the deep learning of LIM.

The Pusa landslide, in Guizhou, China, occurred on 28 August 2017, caused 26 deaths with 9 missing. However, few studies about the pre-event surface deformation are provided because of the complex landslide formation and failure mechanism. To retrieve the precursory signal of this landslide, we recovered pre-event deformation with multi-sensor synthetic aperture radar (SAR) imagery. First, we delineated the boundary and source area of the Pusa landslide based on the coherence and SAR intensity maps. Second, we detected the line-of-sight (LOS) deformation rate and time series before the Pusa landslide with ALOS/PALSAR-2 and Sentinel-1A/B SAR imagery data, where we found that the onset of the deformation is four months before landslide event. Finally, we conceptualized the failure mechanism of the Pusa landslide as the joint effects of rainfall and mining activity. This research provides new insights into the failure mechanism and early warning of rock avalanches.

The extraction of underground minerals in hilly regions is highly susceptible to landslides, which requires the application of InSAR techniques to monitor the surface displacement. However, repeated mining for multiple coal seams can cause a large displacement beyond the detectable gradient of the traditional InSAR technique, making it difficult to explore the relationship between landslides and subsurface excavations in both temporal and spatial domains. In this study, the Tengqing landslide in Shuicheng, Guizhou, China, was chosen as the study area. Firstly, the large-gradient surface displacement in the line of sight was obtained by the fusion of SAR offset tracking and interferometric phase. Subsequently, a multi-segment logistic model was proposed to simulate the temporal effect induced by repeated mining activities. Next, a simplified probability integral method (SPIM) was utilized to invert the geometry of the mining tunnel and separate the displacement of the mining subsidence and landslide. Finally, the subsurface mining parameters and in situ investigation were carried out to assess the impact of mining and precipitation on the kinematic process of Tengqing landslides. Results showed that the repeated mining activities in Tengqing can not only cause land subsidence and rock avalanches at the top of the mountain, but also accelerate the landslide displacement. The technical approach presented in this study can provide new insights for monitoring and modeling the effects of repeated mining-induced landslides in mountainous areas.

Catastrophic landslides occur frequently in Guizhou Province, China, and the landslides in this area have special geomorphological, geological, and anthropogenic features. In order to detect and explore the distribution pattern and control factors of active landslides in Guizhou, firstly, a total of 693 active landslides throughout Guizhou Province were mapped based on the deformation rate, which was obtained by spatiotemporal filtering and Intermittent Small Baseline Subset (ISBAS) Interferometric Synthetic Aperture Radar (InSAR) techniques. Then, the relationships between the detected landslides and elevation, aspect, slope gradient, and stratigraphic lithology were analysed. Moreover, it was found that the landslides were mainly concentrated in three stratigraphic combinations, that is $T_{1}f~P_{2}l-d$, $T_{1}f~T_{1}yn$, and $T_{2}g~T_{1}yn$. Thereafter, the correlation coefficients between the landslide density and elevation and distance to the stratigraphic boundary were 0.54 and −0.19, indicating that the distribution of landslides was significantly controlled by the elevation and the boundary of specific stratigraphic combinations. Finally, we chose a typical landslide to explore how landslide development was controlled by the combined effects of elevation and stratigraphy by using ascending and descending InSAR results. We revealed that landslides occurred primarily in areas with a steep slope and a stratigraphy characterized by mudstone and sandstone.

Landslides pose a significant hazard to lives and property worldwide. Understanding the triggering factors of landslides provides essential information for hazard mitigation. While much research has focused on the effects of precipitation, underground mining, water level changes, and earthquakes on landslides, there remains a gap in understanding the impact of long-term and subtle tectonic interseismic motion, particularly over large-scale areas. Interferometric Synthetic Aperture Radar (InSAR) is widely used in landslide research, effectively detecting wide-area landslides and monitoring high-risk individual landslides. Additionally, it provides insights into the triggering factors and failure mechanisms of landslides. This study focuses on the Chuandian block area in southeastern Tibet, China, an area characterized by active tectonic motion.First, we proposed an automated method for detecting landslides from wide-area InSAR deformation rates, utilizing density clustering and minimum boundary extraction. Using this method, potential landslides were successfully detected in the Chuandian block. The relationship between landslide distribution and the shallow coupling and creep of faults in the Chuandian block was then comprehensively analyzed based on the results of wide-area landslide distribution and interseismic deformation. Specifically, three-dimensional deformation along the Ganzi-Yushu and Xianshuihe faults was monitored using multi-orbit Sentinel-1 SAR and GNSS observations. An elastic dislocation model was also applied to invert shallow creep along these faults. Finally, the development patterns of landslides under the combined influence of internal and external dynamics were summarized. In high-creep areas along the faults, long-term and subtle interseismic motion of the shallow surface led to significant fissure development and structural deterioration in rock and soil, creating internal conditions conducive to landslide formation. External dynamics, including river erosion, precipitation, freezing, and thawing, further accelerated landslide development. Our findings underscore the importance of understanding the relationship between interseismic motion and landslides to enhance knowledge of how tectonic processes influence landslide formation and to support improved hazard mitigation strategies.